Trump addresses AI summit amid energy concerns

On July 15, 2025, President Donald Trump attended a significant summit focused on Artificial Intelligence and its related energy infrastructure at Carnegie Mellon University in Pittsburgh, Pennsylvania. The summit emerged as a crucial platform for discussing the future of AI technologies as momentum builds towards their international governance. During the event, major investments were announced by Google and Blackstone Capital to develop data centers and hydroelectric dams intended to support the growing AI economy in the region. However, the discussions notably lacked a comprehensive exploration of the material requirements necessary for the development and scaling of AI technologies. As documented by Kate Crawford, a notable Microsoft researcher and author of the book Atlas of AI, the material demands that underpin AI technologies, while critical, have been grossly underrepresented in contemporary research and discourse. The focus has predominantly been on the energy intensity and carbon emissions associated with AI rather than the raw materials required to construct the underlying infrastructure. For instance, previous reports highlighted the substantial copper demand linked to the electricity infrastructure of data centers, but only a fraction of the analysis addressed the materials needed for the data centers themselves, which have an annual requirement of approximately 200,000 metric tons. Additionally, discussions surrounding technology innovations, particularly concerning AI chips, have recently pivoted toward gallium due to its potential for enhancing performance. However, comprehensive forecasts concerning its supply have remained limited. Other essential materials, like indium and arsenic, are also important for advanced chip technologies but, similar to gallium, have uncertain future demand predictions. Notably, high purity alumina, which presents avenues for innovation from existing material stocks, is to be sourced primarily from Australia, which is constructing the world's largest factory for its production in Gladstone, Queensland. Looking ahead, quantum computing may play a vital role in supplementing traditional AI processors, heralding a shift toward new materials such as boron and ytterbium, which offer potential advantages for quantum technology applications. The continuing evolution of superconductors with their distinctive material needs underscores the significance of addressing these aspects in parallel with planetary energy and environmental considerations. As international governance strategies for AI are pondered, there is an urgent call for detailed assessments of mineral demand linked to specific national targets for AI infrastructure and for finding ways in which AI can itself contribute to material efficiency and reduce resource constraints.